RMP.11:10 - Rationale
A dated sequence, support for a causal explanation and authority for a present choice have different grounds. A chronology can be established while attribution remains open. A non-causal observation can still identify a design constraint.
Variation, transmission, fashion, retention and loss can suggest hypotheses about Method change. Compare the observations each hypothesis predicts using process-tracing methods. Apply C.28 to identify the evidence required for a causal use, including identification or estimation when that particular question calls for it.